"Artificial intelligence system for product quality measurement"
Our product quality measurement system "FLEXQUALITY" is based on a branch of artificial intelligence called "machine learning", specifically in a technique of deep learning.
It uses "artificial neural networks" which are computational models that process information (in our case images obtained from the calibrator cameras correctly segmented, creating multi-images) imitating the mechanism of human biological neurons, that is, imitating human behavior in image recognition.
In summary, the goal is to enable computer systems to function similarly to the human brain in terms of learning and thinking. However, with clear advantages such as processing speed, consistency, and the lack of degradation in results caused by factors like fatigue or loss of attention.
The machine vision algorithm will create an image of the same fruit in different positions based on repetition, as many positions as the configuration of the camera (or cameras in case of multi-camera systems) allows depending on the synchronization with the calibrator and its speed, this creates something called multimage, which will be used to train the neural network to detect a particular defect and in to infer the process to obtain the classification by quality:
By training a neural network with images captured in real time from the cameras installed in the machine vision area of the sorter, and the subsequent segmentation and treatment creating the multi-images, it will be possible to detect different defects in the product and thus classify by product quality.
These fruit/vegetable multi-images are created in order to obtain an image with as much product texture as possible. Many multimages of the product must be captured to train the neural network with as much of the captured texture of the fruit as possible, in order to determine the degree of damage of any disease or defect that is visible in the product.
Our system is the most advanced in the market for quality determination by artificial intelligence and with the best cost/benefit ratio.
